Integration or Segregation: How Do Molecules
Behave at Oil/Water Interfaces?
F. G. MOORE† AND G. L. RICHMOND*,‡
†
Department of Physics, Whitman College, Walla Walla, Washington 99362,
‡
Department of Chemistry, University of Oregon, Eugene, Oregon 97403
RECEIVED ON DECEMBER 11, 2007
CON SPECTUS
I
t has been over 250 years since Benjamin Franklin, fascinated with the wave-stilling effect of oil on water, performed his famous oil-drop experiments; nevertheless, the
behavior of water molecules adjacent to hydrophobic surfaces continues to fascinate today. In the 18th century, the
calming of the seas seemed the most pertinent application of
such knowledge; today, we understand that oil-on-water phenomena underlie a range of important chemical, physical,
and biological processes, including micelle and membrane
formation, protein folding, chemical separation, oil extraction, nanoparticle formation, and interfacial polymerization.
Beyond classical experiments of the oil-water interface,
recent interest has focused on deriving a molecular-level picture of this interface or, more generally, of water molecules
positioned next to any hydrophobic surface. This Account
summarizes more than a decade’s work from our laboratories aimed at understanding the nature of the hydrogen
bonding occurring between water and a series of organic liquids in contact. Although the common perception is that water molecules and oil molecules positioned at the interface between
the immiscible liquids want nothing to do with one another, we have found that weak interactions between these hydrophilic and hydrophobic molecules lead to interesting interfacial behavior, including highly oriented water molecules and layering of the organic medium that extends several molecular layers deep into the bulk organic liquid. For some organic liquids,
penetration of oriented water into the organic layer is also apparent, facilitated by molecular interactions established at the
molecularly thin region of first contact between the two liquids. The studies involve a combined experimental and computational approach. The primary experimental tool that we have used is vibrational sum frequency spectroscopy (VSFS), a
powerful surface-specific vibrational spectroscopic method for measuring the molecular structures of aqueous surfaces. We
have compared the results of these spectroscopic studies with our calculated VSF spectra derived from population densities and orientational distributions determined through molecular dynamics (MD) simulations. This combination of experiment and theory provides a powerful opportunity to advance our understanding of molecular processes at aqueous interfaces
while also allowing us to test the validity of various molecular models commonly used to describe molecular structure and
interactions at such interfaces.
Introduction
inured to these occurrences have we become that
it is difficult to realize even these most basic oil-
We are all familiar with the adage that “oil and
on-water behaviors mesmerized accomplished
water don’t mix”. Likely too, we are all familiar
naturalists 200 years ago. Indeed, Benjamin Fran-
with the practical consequences of this reality from
klin found these effects so intriguing that he
oil slicks to prismatic layers on rain puddles. So
vowed to carry oil with him “in the hollow joint of
Published on Web 05/29/2008
www.pubs.acs.org/acr
10.1021/ar7002732 CCC: $40.75
Vol. xxx, No. xx
© XXXX American Chemical Society
Month XXXX
000
ACCOUNTS OF CHEMICAL RESEARCH
A
Integration or Segregation Moore and Richmond
my bamboo cane” whenever he went into the country so as
to be able to repeat his experiment: the creation of acre-wide
oil slicks on ponds using mere teaspoons of oil.
The questions he may well have puzzled over while
observing those slicks are still very relevant to us today. What
interactions exist between oil and water as the oil spreads on
a water surface? Do these interactions result in a very abrupt
interface between the two substances or is it more gradual?
Are there different kinds of oil (or other substances) that
exhibit these behaviors to a greater or lesser degree? How do
ions approach and transport across this interface? What might
be learned from this interface that will help us understand
water at other insoluble fluid surfaces or membranes?
In more modern eras, understanding these issues has implications well beyond mere curiosity. Ion transport across membranes, protein folding, chemical separation strategies, and
nanoparticle formation are only a few of the multitude of processes that involve the interaction of a polar water phase with
a hydrophobic oil/organic phase.
In this paper, we provide a summary of the insights gained
in recent years in our laboratory about how interfacial molecules structure and bond at oil-water interfaces. The picture
that has evolved is the result of a closely coupled experimental and computational effort. Experimentally, surface vibrational sum frequency spectroscopy (VSFS) is employed to
measure the vibrational spectrum of interfacial water molecules, and from this we derive an initial picture of how the
organic and water molecules structure, orient, and bond at the
interface. Concurrently, molecular dynamic calculations are
used to provide additional details about the bonding, population density, and orientation of interfacial species and their
VSF response. Knowing the individual VSF response from each
of the contributing species, we then calculate an overall VSF
spectrum that, by comparison with experiment, provides a
check on the models used in the simulations. This combined
approach provides a richer picture of the molecular behavior
at these complex interfaces.
Experimental Considerations
VSFS is a nonlinear optical method that allows the measurement of the vibrational spectrum of molecules at surfaces.1,2
Because it is a second-order process, it is only allowed in
regions of the medium where symmetry is broken, such as at
an interface.3 Figure 1 shows a cartoon diagram for a typical
experimental cell configuration. In the experiment, a visible
laser beam (ωVIS) and a tunable infrared (IR) laser beam (ωIR)
are coincident at the interface. By scanning the energy of the
IR laser over the vibrational mode energies of the surface molB
ACCOUNTS OF CHEMICAL RESEARCH
000
Month XXXX
Vol. xxx, No. xx
FIGURE 1. Experimental geometry for VSF spectra of a liquid/liquid
system studied in TIR geometry.
ecules and monitoring the generated sum frequency signal at
ωSF ) ωIR + ωvis, a vibrational spectrum of the interfacial molecules is obtained. The intensity of the sum-frequency radiation is proportional to the intensity of each incident beam and
(2)
to the square of the effective second-order susceptibility, χeff
:
(2) 2
I(ωSF) ∝ |χeff
| I(ωvis)I(ωIR)
(1)
(2)
χeff
can be expressed as the sum of resonant and nonreso(2)
(2)
nant components, χ(2)
R and χNR. Where χNR is usually treated as
a constant to be determined in the experiment. The resonant
component is frequency dependent and proportional to the
number density of molecules, N, and the orientationally averaged molecular hyperpolarizability, βv, as follows
χR(2)v )
N
〈β 〉
ε0 v
(2)
The molecular hyperpolarizability, βv, is enhanced when
the frequency of the IR field is resonant with a SF-active vibrational mode of a molecule at the surface or interface. This
enhancement in βv leads to an enhancement in the nonlinear susceptibility, χR(2)v , which contains the product of the
Raman and the IR transition moments.
This spectroscopic approach (VSFS) has advantages over linear spectroscopies, because it is highly specific to the thin
layer of molecules within the interfacial region, this region
itself being defined by the extent of anisotropic orientation
and bonding of its molecules relative to the bulk phase. The
disadvantages are more extensive instrumentation and the
more complex spectral response. Since VSFS is a coherent
spectroscopy, each oscillator’s contribution to a response has
an associated phase that can interfere constructively or
destructively with adjacent modes, making analysis and spectral assignments challenging. This coherence can also be an
advantage because, when interpreted appropriately, it can provide information about molecular orientation in the interfacial region.
Integration or Segregation Moore and Richmond
The experimental sum-frequency generation (SFG) response
depends on the polarization of each of the two incident beams
and the sum-frequency emission generated is itself polarized.
By adjusting polarization combinations of incident and outgoing light, the different SFG responses (each related to one
or more elements of the χ(2) tensor) can be selectively probed.
The notation for these sorts of studies requires distinguishing between light polarized parallel to the plane of incidence
(P) and perpendicular to the plane of incidence (S). All polarization notations are given in the order sum-frequency, visible, IR, for example, (SPS).
All of the liquid/liquid studies described herein were
accomplished on Nd:YAG laser based systems with the IR
(from nonlinear optical generation methods) and visible beams
incident on the interface through a prism in a total internal
reflection (TIR) geometry to maximize the weak VSF signal.
The details of the experimental systems used to collect the
spectra shown in this Account have been discussed in previous publications.4–7 The cells are typically made of Kel-F to
withstand the requisite aggressive cleaning protocols and contact with the organic liquids used. For our chloroform/water
investigations, deuterated chloroform (CDCl3) was used. This
was necessitated by the comparatively strong IR absorption
that occurs in CHCl3.
Insights from Spectroscopic Measurements
Our VSF studies of oil-water interfaces have provided a progressively deeper understanding of the interactions at the
interface between water and several hydrophobic organic liquids and films of varying polarity and molecular structure. Figure 2 shows the VSF spectra of interfacial water for three of
these systems6–9 along with the air/water interface10,11 for
comparison. The spectra reveal OH stretch modes of water
molecules that are highly sensitive in frequency and intensity to the degree of hydrogen bonding between molecules.12,13 The polarizations employed here (SSP) probe OH
modes that have a dipole component perpendicular to the surface plane.
The CCl4/H2O, CDCl3/H2O, and vapor/H2O interfaces show
similar spectral features, indicative of general similarities in the
orientation and hydrogen bonding of water at these interfaces.
The often-found peak near 3700 cm-1, is referred to as the
“free OH” mode. This peak arises when one of the OH bonds
of a water molecule orients into the hydrophobic phase,
whether that be organic liquid or air. Its companion OH bond
(referred to commonly as the “donor” mode because it can act
as a hydrogen bond donor to nearby water molecules) orients into the aqueous phase and is one of the primary con-
FIGURE 2. Resonant VSF spectra of water at three organic/water
interfaces and the air/water interface. The dotted line through the
cartoon molecules represents the interface, and the associated
arrow indicates the free OH bond. The red line is a guide to the eye
for the free OH frequency. The air/water spectrum (gray) contains a
large nonresonant (NR) background. With its NR background
removed (blue) a comparison with organic/water systems (having
almost zero NR background) is made possible. Spectra are taken
with SSP polarization.
tributors to intensity in the 3200-3500 cm-1 region. Its VSF
contribution is red-shifted and broader than the free OH mode
because it is bonded to nearby water molecules. There are
also contributions in the 3000-3500 cm-1 region from stronger bonded water species, but these are nearly impossible to
characterize based on any spectral fitting routines. We have
used two methods for deconvoluting these spectral regions,
spectral fitting with isotopic dilution experiments using HOD/
D2O mixtures6,14 and molecular dynamics simulations that
allow extraction of different water bonded species based on
their calculated VSF response.15–17 The former method works
well for understanding weakly bonded and vibrationally
uncoupled interfacial water species, and the latter assists in
further understanding of both weakly and more strongly
bonded interfacial water species.
There are a number of important general insights gained
from the systems studied thus far. First, the presence of oriented water molecules that have a free OH mode indicates
that the interfacial region for most of the systems that we have
studied is relatively sharp, certainly within the dimensions of
5-10 Å, consistent with X-ray data of similar liquid/liquid
Vol. xxx, No. xx
Month XXXX
000
ACCOUNTS OF CHEMICAL RESEARCH
C
Integration or Segregation Moore and Richmond
perfluorinated and hydrogenated monolayers, which are discussed further below.
For several of the organic/water systems, we find spectral
intensity near the free OH resonance, which suggests the presence of water molecules that have negligible interaction with
other water molecules but that are highly oriented with their
dipoles perpendicular to the interface and their hydrogens
pointing into the organic phase. These “water monomers”
bond weakly to neighboring interfacial organic molecules or
as OH bond acceptors to other water molecules, a factor we
believe contributes to the overall orientation of the interfacial region.6 They appeared in our earliest work with CCl4/
H2O and our latest work with chloroform. Our ability to see
the small number of these weakly bonded water molecules is
largely due to the windfall of spectral interferences that come
with the VSF technique. These will be discussed in more detail
in the molecular dynamics (MD) section.
Interestingly, the neat CCl4/H2O spectrum, which we have
FIGURE 3. Resonant VSF spectra of water at the interface of three
alkane/water systems. Taken with SSP polarization. Data from ref 9.
TABLE 1. VSFS Measured Free OH Frequency of Water Measured at
Different Interfaces
system
free OH (cm-1)
vapor/H2O
F-monolayer/H2O
H-monolayer/H2O
hexane, heptane, octane
CCl4/H2O
CDCl3/H2O
3705 ( 5
3694 ( 3
3674 ( 2
3674 ( 3
3669 ( 2
3650 ( 3
interfaces.18–20 Second, we have found that the free OH frequency is an indicator of the presence of weak water-organic
interactions present at these interfaces and the relative
strength of this interaction.21 For example, for CCl4/H2O, the
free OH appears at 3669 ( 4 cm-1,22 red-shifted from its
value at the vapor/water interface at 3705 cm-1.14 The larger
red shift for chloroform (∼3650 cm-1) is indicative of a stronger interaction between water and this slightly more polar
molecule.7
The alkane/water interfaces show spectral responses similar to those found with CCl4/H2O and CDCl3/H2O (Figure 3).
However the free OH is at 3674 cm-1,23 indicative of a
weaker alkane-water interaction than that for the chlorinated
solvents. We conclude that the broader free OH peak found
for all the alkanes represents a more diverse range of
water-alkane interfacial interactions. Interestingly, all alkanes
studied give the same free OH frequency. Table 1 summarizes the free OH frequency of systems studied in this laboratory. F-monolayer and H-monolayer entries refer to
D
ACCOUNTS OF CHEMICAL RESEARCH
000
Month XXXX
Vol. xxx, No. xx
studied most extensively, is remarkably similar to the resonant vapor/water interfacial spectrum. The vapor/water spectrum shown in Figure 2 (dotted line) contains a significant
nonresonant background, which is not present for the organic/
water systems studied. When this nonresonant contribution is
removed using isotopic dilution experiments, the resonant
vapor/water spectrum (shown in blue) shows even stronger
resemblance to the organic/water interfaces.
Of the systems discussed thus far, the spectrum of the 1,2dichloroethane (DCE)/water interface shown in Figure 2 is strikingly different in displaying no free OH mode. The DCE/H2O
interface has been the subject of considerable attention in the
electrochemistry community because of its relatively large
polarizable window and the question as to its nature being
either sharp or diffuse.24 Given the observed lack of distinct
spectral features found for DCE/H2O, we opted to explore this
system further by “doping” increasing amounts of DCE into the
organic phase of the more spectrally distinct CCl4/H2O
system.7,17 As shown in Figure 4a, with increased amounts of
DCE added, the free OH from neat CCl4 red-shifts and the free
OH oscillator intensity decreases significantly. This leads us to
conclude, as we initially did, that the interface for the DCE/
H2O system is much more diffuse, with a lack of distinct water
orientation found in the other systems. Here, insights from our
MD simulations have been invaluable. Our simulations readily
reproduced the spectral trends as shown in Figure 4b. When
we extract the bonding and orientation information from these
simulations, which so well fit the data, a very different and fascinating picture emerges and will be described later.
Integration or Segregation Moore and Richmond
spectral profile and the free OH behavior mirrors that of the
alkane/water systems.26 The small red shift in the free OH frequency for the fluorocarbon monolayer system relative to air/
water shows that a weak water-fluorocarbon interaction is
present at these interfaces.
Insights from Molecular Dynamics
Simulations
FIGURE 4. Experimental (a) and computed (b) spectra of the (CCl4
+ DCE)/H2O interface where the mole fraction of DCE ranges from
zero (top) to 1.0. Intermediate concentrations are approximately
8%, 12%, 20%, and 35% DCE. Spectra have been shifted vertically
for clarity. A dashed black line has been added to help identify the
small red shift in the total free OH peak. Reproduced from ref 17.
Copyright 2007 American Chemical Society.
Additional studies have probed how water at hydrophobic monolayer/water interfaces compare with these liquid/liquid interfaces. The results show that water interacts at the
terminal end of these organic monolayers in a very similar
manner as at the liquid/liquid interfaces.25 Two types of
monolayers have been examined under water, hydrogenated
(octadecyltrisilane) and perfluorinated monolayers (1H,1H,2H,2H-perfluorodecyl-siloxane, FTDS) bound to Si/SiO2.26 For
well-formed monolayers, we find that a free OH is present for
water at the terminal portion of both the hydrocarbon and fluorocarbon monolayers. For hydrocarbon monolayers, the OH
Our MD work is used to complement and enhance our experimental studies and falls into two broad categories. The first
uses and expands upon a formalism introduced by Hirose et
al.27 in which MD simulations are used along with hyperpolarizability calculations to compute the VSF response28 of interfacial water for the air/water and organic/water systems we
have studied.15,17,29 With good correlation between calculated and experimental spectra, we proceed to extract detailed
information from the simulations about the different water
bonded species in the interfacial region, their degree of bonding, orientation, and interfacial depth. The second MD effort
examines the orientation and ordering of both the water and
“oil” species in the interfacial region.30,31
The starting point for all of this work is fully atomistic,
polarizable descriptions of the molecules under study. These
descriptions are combined with initial “boxes” of molecules
chosen to accurately reflect the actual bulk experimental densities of the systems being simulated. Volumes of the simulated systems are on the order of 13 000 Å3, and they
typically contain on the order of 2000 water molecules and
between 400 and 500 “oil” molecules in order to get the densities correct. The simulation environment used for this work
has changed over the years; most recently the AMBER7 suite
of programs has been utilized.32
Our MD simulations of the various liquid/liquid interfaces
discussed above all display a high degree of water orientation in the interfacial region. For water and carbon tetrachloride, weak bonding interactions play a significant role in
establishing interfacial orientation (see below for further examination of this). For the CCl4/water, chloroform/water, and
alkane/water systems, we find that the interfaces are molecularly sharp with widths of ∼5-6 Å, which correlate well with
simulations of others on related systems.33,34 The calculations confirm our experimentally derived conclusions that a
small but significant number of water molecules are found to
interact weakly with the alkanes and CCl4 despite their nonpolar nature. For the alkanes, two types of weak interactions
with water are present and are responsible for the observed
broadening of the free OH peak relative to CCl4/H2O and
CDCl3/H2O.23 The more polar organics studied in the simulaVol. xxx, No. xx
Month XXXX
000
ACCOUNTS OF CHEMICAL RESEARCH
E
Integration or Segregation Moore and Richmond
FIGURE 5. Picture of the DCE/water (right) and CCl4/water (left) interfaces derived from the MD simulations.
tions, DCE and dichloromethane, also form well-defined interfaces. Figure 5 displays pictures for visual comparison from
the simulations of the CCl4/H2O and the slightly more diffuse
DCE/H2O interfaces. Orientational and hydrogen-bonding
analysis for all three systems show a large density of water
molecules in the interfacial region that orient with their dipoles
almost parallel to the interfacial plane and significant bonding with adjacent water molecules. This population of nearly
in-plane water molecules is “invisible” to VSF in any polarization combination because of their isotropic in-plane orientation. The simulations for all systems also show to varying
degrees a small number of very weakly interacting water molecules that are in the organic-rich interfacial region and are
oriented with their hydrogens directed toward the organic bulk
phase, as has been suggested by our experiments.
All interfaces are awash with water molecules that exhibit
a constellation of different bonding interactions and orientations. Our calculations allow us to probe this further as we
determine the number density for the different mono-, di-, tri-,
and tetracoordinated water species at the interface, their polar
orientation, and the VSF contribution that each type of bonded
molecule makes to the overall measured spectrum. In characterizing these interactions, we have used the following naming scheme (depicted in Figure 6):35 Tetracoordinated
molecules are denoted as OOHH-bonded species, tricoordinated species can be either OOH- or OHH-bonded, dicoordinated are OH-, HH-, or OO-bonded species, and monocoordinated species are designated as H- or O-bonded. Water
molecules that straddle the interface have a free OH bond and
a companion donor OH bond that can participate in H, OH, or
OOH bonding.
As an example of the information extracted from these
simulations, Figure 7a shows the results for the populations of
these different water bonded species across the interfacial
F
ACCOUNTS OF CHEMICAL RESEARCH
000
Month XXXX
Vol. xxx, No. xx
FIGURE 6. Different contributing water bonded species and the
labeling scheme. The blue shading indicates whether the atom is
involved in H bonding to nearby water molecules. Dotted green
circles indicate which H atom is part of the “free OH” oscillator.
region for the CCl4/water interface. Peaking in the center of
the interfacial region are water molecules that straddle the
interface as evidenced by the relatively high population of free
OH and donor OH (with OH bonding) near the center of the
interfacial region. These two modes are also found in related
calculations to be highly oriented with each showing a preferred orientation perpendicular to the surface but antiparallel to one another, consistent with the straddling water
molecule picture. Tetrahedrally bonded water and water molecules with OHH bonding are also evident but tend to show
their presence deeper in the interfacial region and, of course,
exist into the bulk aqueous phase. Within the significant population of topmost water molecules that orient with their
Integration or Segregation Moore and Richmond
FIGURE 7. (a) Density profiles across the interface for the dominant
water species at the CCl4/H2O interface. Labels refer to the bonded
species in Figure 6. (b) Calculated VSF spectrum of the CCl4/H2O
interface (red) and the most dominant contributors to the SSP
spectrum. Adapted from ref 17.
dipoles largely in the plane of the surface, these tend to be
involved in OHH bonding.
The conclusions drawn from these calculations can be validated by incorporating these density profile plots and related
orientational plots into a calculation of a VSF spectrum for
these systems. Comparison of the calculations with the experimental results (See Figure 4, for example) show that for the
CCl4/H2O and DCE/H2O interfaces there is remarkable agreement. Although there is still room for improvement, that such
agreement can be achieved even with the rather stark differences in polarity and symmetry of the nonaqueous components is a testament to the rigor and robustness of the
approach.
The calculations provide additional information about what
species make the largest contribution to the VSF spectrum for
the most commonly used SSP polarization.17 The results for
the dominant species for CCl4/H2O are depicted in Figure 7b.
As shown, most of the intensity for this polarization comes
from those molecules that are highly oriented and that straddle the interface. The greatest number of other contributions
come from OOH- and OHH-bonded water molecules. A small
number of oriented water monomers (no bonding to water)
are present in the more organic-rich interfacial region and orient with their hydrogens directed into the CCl4 phase. The
contribution from these monomers is negligibly small because
of their low densities and is included in the “all others” category. These results confirm our experimental observations and
will be discussed in more detail below.
Along with our liquid/liquid effort, we have conducted similar experimental and computational studies of the vapor/water interface,15 following computationally on the work of
Morita and Hynes.28 It is important to note that we find that
the CCl4/water interface shows striking similarity to the vapor/
water interface except for the free OH frequency and the presence of unbonded and oriented monomers. This is consistent
with theoretical studies36 that suggest weak bonding interactions at a water/hydrophobic surface although our studies do
not provide evidence for any significant vacuum-like drying
region at this interface.
We now return to the question of why the DCE/H2O spectrum is so different from the alkane/water, CCl4/H2O, and the
CDCl3/H2O spectra. Is the spectrum telling us that the DCE/
H2O interface is truly broad and disordered as a cursory
examination of the featureless spectra might suggest? The MD
simulations provide needed insight.37 The results show that
the interfacial population density and interfacial orientation of
water molecules near DCE is quite similar to the other systems except for several notable variations that result in the
decidedly different VSF spectrum. These are best demonstrated visually from our depth spectral profiling results shown
in Figure 8. In these studies, we have calculated the spectral
profile of the different water bonded species as a function of
interfacial depth (+z for the water region and -z for the
organic).16,37 Since the generation of VSF intensity comes from
water molecules that have an anisotropic distribution in the
medium, the z-dimension breadth of the region from which
the intensity originates is a good measure of the depth of the
interfacial region on both sides of the interface. Figure 8 (top)
shows the total SSP VSF response plotted as a function of
interfacial depth for CCl4/H2O (left) and DCE/water (right). In
contrast to CCl4, with the slightly more polar DCE, the greater
width of the spectral profile for the later indicates that this
interface is indeed more diffuse than the CCl4/H2O interface,
consistent with other MD studies.34 The free OH is clearly
present at the DCE/water interface (Figure 8 center) but covers a broader spatial and spectral region due to stronger and
more diverse water-organic interactions. These effects conspire to produce a combination of spectral congestion and
interferences that result in a largely featureless spectrum.
Related MD simulations focus specifically on molecular orientation and ordering in the interfacial region. Since the simulations are of equilibrium conditions, the coordinate sets can
be analyzed for statistical information about the location and
Vol. xxx, No. xx
Month XXXX
000
ACCOUNTS OF CHEMICAL RESEARCH
G
Integration or Segregation Moore and Richmond
FIGURE 8. Computational VSF spectral profiles as functions of frequency (horizontal axis) and interfacial depth (vertical axis) in SSP for
CCl4-H2O (left) and DCE-H2O (right) interfaces. Resolution (bin size) for each graph is 1 Å and 1 cm-1: (top) VSF depth profile for all
contributing species; (middle) the free OH modes; (bottom) water monomers. Adapted from ref 37.
orientation of molecules within the system.30,31 Typically, the
species of interest is the minority species, for example, oil molecules that have crossed the interface into the water region or
H2O molecules that have moved into the oil region. Particularly for polar molecules, statistics of these distributions are
useful and can be used to generate order parameters for the
molecular configuration. With appropriate visualization techniques, these derived data allow a sophisticated understanding of the interfacial environment.30,31
An exceptionally interesting result is the orientational layering of the organic side of the interface found for the CCl4/
H2O system.30 A study of the CCl4-vapor interface has found
no preferred orientations of the surface CCl4 molecules,38
alluding to the role that water likely plays in supporting this
layered structure. The cartoon in Figure 9 indicates the sort of
layering arrangement that can be discerned by an orientational analysis of MD data.31 That the molecules in the sysH
ACCOUNTS OF CHEMICAL RESEARCH
000
Month XXXX
Vol. xxx, No. xx
tem are oriented and interacting in this fashion is certainly
taken into account during the calculation of a simulated VSF
spectrum, but the ability to visualize the structure of the
organic provides for a parallel understanding and appreciation of the microstructure experienced by the constellation of
H2O oscillators present in this region of the interface. We
attribute the alignment in part to the presence of polar oriented interfacial water molecules interacting with the polarizable CCl4.
Another oil, chloroform, has a uniquely defined directional
aspect (the C-H bond direction), and both the population density statistics and orientation statistics for this system have
been studied.31 The take-home message from this study is
that the average chloroform molecule in the interfacial region
is oriented such that its C-H bond is directed toward the H2O
phase. This is not the entire story, however, because it is also
the case that studies indicate that a layered ordering exists in
Integration or Segregation Moore and Richmond
North America with a supply of oil concealed in his walking
cane. The industrial revolution has come and gone, and the
21st century has opened with the promise of great technical
and scientific advances but also with an ever-increasing
awareness and concern as to the environmental impact of global development. The curious behavior of the oil/water interface is no different today than it was in Franklin’s time.
However, interest in understanding these layers has surely
morphed into something far beyond simple curiosity.
In these studies, we have found that at the closest contact
point between water and the organic, the interfacial water
shows very weak bonding character with other water molecules, similar to what is observed at the vapor/water interface. We have also learned about the important role that weak
water-organic interactions have in the structuring of the interFIGURE 9. At an aqueous interface, carbon tetrachloride molecules
form an alternating layer structure, consisting of corner and face
contacts between tetrahedra. The width of most of the layers
corresponds to that of a single CCl4 molecule. Reproduced from ref
30. Copyright 2007 American Chemical Society.
the C-H bond statistics. While the population of chloroform
molecules decreases monotonically with distance into the H2O
phase, there are a series of alternating orientations assumed
by the molecules making up this distribution such that their
hydrogens are alternately pointing deeper into the water
phase or out of the water phase.
The picture is more complicated for H2O near the chloroform/water interface because its molecular structure is such
that it has not one but two orientational angles to be considered. Hence, there are two order parameters associated with
water in the chloroform/water system, one for water tilt and
one for water twist. Visualization of these statistics is significantly more challenging and is best considered in detail in the
original publication.31 To appreciate a summary of this work,
one must understand that ordering in the chloroform/water
system occurs in a balance between dipole compensation and
the optimization of the hydrogen-bonding network. Portions
of the interfacial region may be more favorable to a particular hydrogen-bonding geometry, and in these there may be a
net dipole orientation resulting in a net volume charge density on the water side of the interface. We have suggested that
such a net orientation could help explain the presence of an
enhanced hydroxide ion concentration at hydrophobic/water
surfaces.
face. Across all the oil/water systems that we have studied,
these weak interactions drive a molecular ordering behavior
that can extend well into the organic phase. The resulting net
orientation creates fields that may assist in the incursion of
polar species into the organic phase, as well as serving to
draw ions into the interfacial region. Both of these topics are
currently under further investigation in our laboratory.
Key to our nuanced understanding is the application of an
appropriate surface-specific experimental probe of molecular
structure and orientation. When combined with molecular
dynamics simulations that yield both a calculated nonlinear
optical response and molecular orientation, we can achieve a
triangulation of sorts that unerringly points toward the actual
microstructure at the interface. This combined approach, as
well as exciting new advances from other research efforts in
the application of X-ray and neutron scattering techniques and
other simulation efforts, is surely helping us answer many of
the questions that Benjamin Franklin would have wanted to
ask about the oil/water interface. There are many frontiers
ahead in this area in both experiment and theory. The understanding of these interfaces will continue to evolve with
improvements in probe techniques and improved molecular
models for simulating complex interactions. As such, this
review provides a snapshot of where nonlinear spectroscopic
measurements coupled with MD simulations have brought us
in recent years.
This research was supported by funds from the National Sci-
Conclusions and Future Directions
ence Foundation (Grant CHE-0652531). Funding from the
It has been more than 200 years since Benjamin Franklin purportedly wandered the environs of the eastern seaboard of
Office of Naval Research assisted in instrumentation and the
computational studies.
Vol. xxx, No. xx
Month XXXX
000
ACCOUNTS OF CHEMICAL RESEARCH
I
Integration or Segregation Moore and Richmond
BIOGRAPHICAL INFORMATION
Fred Moore’s research interests relate to understanding inhomogeneous systems by applying spectroscopic and modeling
techniques. He received B.A. degrees in mathematics and physics from Lewis & Clark College and a Ph.D. in Applied Physics
from the Oregon Graduate Center. After two years as a National
Research Council Postdoctoral fellow at the Naval Research Laboratory, he joined the faculty of Whitman College where is now
Professor and Chair of Physics.
Geraldine Richmond’s research interests are in understanding
molecular structure, bonding, and dynamics at complex interfaces. She received her B.S. degree at Kansas State University and
her Ph.D. from the University of California at Berkeley under the
guidance of George Pimentel. After five years as an assistant professor at Bryn Mawr College, she moved to the University of Oregon where she currently resides as the Richard M. and Patricia H.
Noyes Professor of Chemistry.
REFERENCES
1 Du, Q.; Superfine, R.; Freysz, E.; Shen, Y. R. Vibrational spectroscopy of water at the
vapor/water interface. Phys. Rev. Lett. 1993, 70, 2313–2316.
2 Shen, Y. R. Surface properties probed by second harmonic generation and sumfrequency generation. Nature 1989, 337, 519–525.
3 Bloembergen, N.; Pershan, P. S. Light waves at the boundary of nonlinear media.
Phys. Rev. 1962, 128, 606–622.
4 Conboy, J. C.; Messmer, M. C.; Richmond, G. L. Dependence of alkyl chain
conformation of simple ionic surfactants on head group functionality as studied
by vibrational sum-frequency spectroscopy. J. Phys. Chem. B 1997, 101,
6724–6733.
5 Conboy, J. C.; Messmer, M. C.; Richmond, G. L. Investigation of surfactant
conformation and order at the liquid-liquid interface by total internal reflection
sum-frequency vibrational spectroscopy. J. Phys. Chem. 1996, 100, 7617–
7622.
6 Scatena, L. F.; Richmond, G. L. Orientation, hydrogen bonding, and penetration
of water at the organic/water interface. J. Phys. Chem. B 2001, 105, 11240–
11250.
7 McFearin, C. L.; Richmond, G. L. Understanding how organic solvent polarity affects
water structure and bonding at halocarbon-water interfaces. J. Mol. Liq. 2007, 136,
221–226.
8 Richmond, G. L. Structure and bonding of molecules at aqueous surfaces. Annu.
Rev. Phys. Chem. 2001, 52, 357–389.
9 Walker, D. S.; Brown, M. G.; McFearin, C. L.; Richmond, G. L. Evidence for a diffuse
interfacial region at the dichlorethane/water interface. J. Phys. Chem. B 2004, 108,
2111–2114.
10 Raymond, E. A.; Tarbuck, T. R.; Brown, M. G.; Richmond, G. L. Hydrogen bonding
interactions at the vapor/water interface as investigated by vibrational sum
frequency spectroscopy of HOD/H2O/D2O mixtures and molecular dynamics
simulations. J. Phys. Chem. B 2003, 107, 546–556.
11 Richmond, G. L. Molecular bonding and interactions at aqueous surfaces as probed
by vibrational sum frequency spectroscopy. Chem. Rev. 2002, 102, 2963–2724.
12 Jeffrey, G. A. An Introduction to Hydrogen Bonding; Oxford University Press: Oxford,
U.K., 1997.
13 Scherer, J. R. The vibrational spectroscopy of water. In Advances in Infrared and
Raman Spectroscopy; Clark, R. J. H., Hester, R. E., Eds.; Heyden: Philadelphia, PA,
1978; Vol. 5, pp 149-216.
14 Raymond, E. A.; Tarbuck, T. R.; Richmond, G. L. Isotopic dilution studies of the
vapor/water interface as investigated by vibrational sum-frequency spectroscopy. J.
Phys. Chem. B 2002, 106, 2817–2820.
J
ACCOUNTS OF CHEMICAL RESEARCH
000
Month XXXX
Vol. xxx, No. xx
15 Walker, D. S.; Richmond, G. L. Understanding hydrogen bonding effects at the
vapor/water interface through molecular dynamics calculations of the vibrational
sum frequency spectra of H2O/D2O/HOD Mixtures. J. Phys. Chem. C 2007, 111,
8321–8330.
16 Walker, D. S.; Richmond, G. L. Depth profiling of water molecules at the liquid-liquid
interface using a combined surface vibrational spectroscopy and moleuclar
dynamics approach. J. Am. Chem. Soc. 2007, 129, 9446–9451.
17 Walker, D. S.; Moore, F. G.; Richmond, G. L. Vibrational sum frequency
spectroscopy and molecular dynamics simulations of the carbon tetrachloridewater and 1,2-dichloroethane-water interfaces. J. Phys. Chem. C 2007, 111 (16),
6103–6112.
18 Luo, G. M.; Malkova, S.; Pingali, S. V.; Schultz, D. G.; Lin, B. H.; Meron, M.; Graber,
T. J.; Gebhardt, R.; Vanysek, P.; Schlossman, M. L. X-Ray studies of the interface
between two polar liquids: neat and with electrolytes. Faraday Discuss. 2005, 129,
23–34.
19 Schlossman, M. L. Liquid-liquid interfaces: Studied by X-ray and neutron scattering.
Curr. Opin. Colloids Surf. 2002, 7 (3-4), 235–243.
20 Mitrinovic, D. M.; Tikhonov, A. M.; Li, M.; Huang, Z. Q.; Schlossman, M. L.
Noncapillary-wave structure at the water-alkane interface. Phys. Rev. Lett. 2000, 85
(3), 582–585.
21 Richmond, G. L. Molecules at aqueous surfaces. Annu. Rev. Phys. Chem. 2001,
292, 257–285.
22 Scatena, L. F.; Richmond, G. L. Water at hydrophobic surfaces: Weak hydrogen
bonding and strong orientation effects. Science 2001, 292, 908–911.
23 Brown, M. G.; Walker, D. S.; Raymond, E. A.; Richmond, G. L. Vibrational sumfrequency spectroscopy of alkane/water interfaces: Experiment and theoretical
simulation. J. Phys. Chem. B 2003, 107 (1), 237–244.
24 Girault, H. H.; Schiffrin, D. J. Thermodynamic surface excess of water and ionic
solvation at the interface between immiscible liquids. J. Electroanal. Chem. 1983,
150, 43–49.
25 Becraft, K. A.; Richmond, G. L. Surfactant adsorption at the salt/water interface:
Comparing the conformation and interfacial water structure for selected surfactants.
J. Phys. Chem. B 2005, 109, 5108–5117.
26 Hopkins, A. J.; McFearin, C. L.; Richmond, G. L. Investigations of the solid-aqueous
interface with vibrational sum frequency spectroscopy. Curr. Opin. Solid State Mater.
Sci. 2005, 9, 19–27.
27 Hirose, C.; Akamatsu, N.; Domen, K. Formulas for the analysis of surface sumfrequency generation spectrum by CH modes of methyl and methylene groups.
J. Chem. Phys. 1992, 96, 997–1004.
28 Morita, A.; Hynes, J. T. A theoretical analysis of the sum frequency generation
spectrum of the water surface. Chem. Phys. 2000, 258, 371.
29 Walker, D. S.; Hore, D. K.; Richmond, G. L. Understanding the population,
coordination, and orientation of water species contributing to the nonlinear optical
spectroscopy of the vapor-water interface through molecular dynamics simulations.
J. Phys. Chem. B 2006, 110, 20451–2059.
30 Hore, D. K.; Walker, D. S.; Richmond, G. L. Layered organic structure at the carbon
tetrachloride-water interface. J. Am. Chem. Soc. 2007, 129 (4), 752–753.
31 Hore, D.; Walker, D. S.; MacKinnon, L.; Richmond, G. L. Molecular structure of the
chloroform-water and dichloromethane-water interfaces. J. Phys. Chem. C 2007,
111, 8832–8842.
32 Case, D. A.; et al. AMBER 7 University of California: San Francisco, 2002.
33 Chang, T.-M.; Dang, L. X. Molecular dynamics simulations of CCl4-H2O liquid-liquid
interface with polarizable potential models. J. Chem. Phys. 1996, 104, 6772–6783.
34 Benjamin, I. Molecular structure and dynamics at liquid-liquid interfaces. Annu. Rev.
Phys. Chem. 1997, (48), 407–451.
35 Buch, V. Molecular structure and OH-stretch spectra of a liquid water surface. J.
Phys. Chem. B 2005, 109, 17771–17774.
36 Chandler, D. Interfaces and the driving force of hydrophobic assembly. Nature 2005,
437, 640–647.
37 Walker, D. S.; Richmond, G. L. Interfacial depth profiling of the orientation and
bonding of water molecules across different organic-water interfaces. J. Phys.
Chem. C 2008, 112, 201–209.
38 Chang, T.-M.; Peterson, K. A.; Dang, L. X. Molecular dynamics simulations of liquid,
interface, and ionic solvation of polarizable carbon tetrachloride. J. Chem. Phys.
1995, 103, 7502–7513.